Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Homophily Enhanced Graph Domain Adaptation
Authors: Ruiyi Fang, Bingheng Li, Jingyu Zhao, Ruizhi Pu, Qiuhao Zeng, Gezheng Xu, Charles Ling, Boyu Wang
ICML 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on a variety of benchmarks verify the effectiveness of our method. (Abstract) and The results of experiments are summarized in Table 1, Table 2 and Table 3, where highest scores are highlighted in bold, and the second-highest scores are underlined. (Section 5.3) |
| Researcher Affiliation | Academia | 1Western University, Ontario, Canada 2Michigan State University, Michigan, USA 3University of Electronic Science and Technology of China, Chengdu, Sichuan Province, China 4Vector Institute, Ontario, Canada. Correspondence to: Boyu Wang <EMAIL>. |
| Pseudocode | Yes | Algorithm 1 The proposed algorithm HGDA (Appendix F) |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. There are no explicit statements about code release or links to a code repository. |
| Open Datasets | Yes | To demonstrate the effectiveness of our approach on domain adaptation node classification tasks, we evaluate it on four types of datasets, including Airport (Ribeiro et al., 2017), Citation (Wu et al., 2020), Social (Liu et al., 2024a), ACM (Shen et al., 2024), and MAG datasets (Wang et al., 2020). (Section 5.1 Datasets) |
| Dataset Splits | No | We repeatedly train and test our model for five times with the same partition of dataset and then report the average of ACC. (Appendix B Experimental Setup) - This sentence implies a partition but does not specify the actual splits (e.g., train/test/validation percentages or counts). |
| Hardware Specification | Yes | The experiments are implemented in the PyTorch platform using an Intel(R) Xeon(R) Silver 4210R CPU @ 2.40GHz, and GeForce RTX A5000 24G GPU. (Appendix B Experimental Setup) |
| Software Dependencies | No | The experiments are implemented in the PyTorch platform using an Intel(R) Xeon(R) Silver 4210R CPU @ 2.40GHz, and GeForce RTX A5000 24G GPU. (Appendix B Experimental Setup) - PyTorch is mentioned, but without a version number. |
| Experiment Setup | Yes | Technically, two layers GCN is built and we train our model by utilizing the Adam (Kingma & Ba, 2015) optimizer with learning rate ranging from 0.0001 to 0.0005. In order to prevent over-fitting, we set the dropout rate to 0.5. In addition, we set weight decay {1e 4, , 5e 3}. For fairness, we use the same parameter settings for all the cross-domain node classification methods in our experiment, except for some special cases. For GCN, UDA-GCN, and JHGDA the GCNs of both the source and target networks contain two hidden layers (L = 2) with structure as 128 16. The dropout rate for each GCN layer is set to 0.3. (Appendix B Experimental Setup) |